Perioperative blood transfusion is not associated with overall survival or time to recurrence after resection of perihilar cholangiocarcinoma

HPB

http://dx.doi.org/10.1016/j.hpb.2015.08.004

ORIGINAL ARTICLE

Perioperative blood transfusion is not associated with overall survival or time to recurrence after resection of perihilar cholangiocarcinoma Annemiek M. Dekker1, Jimme K. Wiggers1, Robert J. Coelen1, Rowan F. van Golen1, Marc G.H. Besselink1, Olivier R.C. Busch1, Joanne Verheij3, Markus W. Hollmann2 & Thomas M. van Gulik1 1

Department of Surgery, 2Department of Anaesthesiology, Academic Medical Centre (AMC), and 3Department of Pathology, Amsterdam Medical Center, Amsterdam, The Netherlands

Abstract Background: Perioperative blood transfusions have been associated with worse oncological outcome in several types of cancer. The objective of this study was to assess the effect of perioperative blood transfusions on time to recurrence and overall survival (OS) in patients who underwent curative-intent resection of perihilar cholangiocarcinoma (PHC). Methods: This retrospective cohort study included consecutive patients with resected PHC between 1992 and 2013 in a specialized center. Patients with 90-day mortality after surgery were excluded. Patients who did and did not receive perioperative blood transfusions were compared using univariable Kaplan–Meier analysis and multivariable Cox regression. Results: Of 145 included patients, 80 (55.2%) received perioperative blood transfusions. The median OS was 49 months for patients without and 41 months for patients with blood transfusions (P = 0.46). In risk-adjusted multivariable Cox regression analysis, blood transfusion was not associated with OS (HR 1.00, 95% CI 0.59–1.68, P = 0.99) or time to recurrence (HR 1.00, 95% CI 0.57–1.78, P = 0.99). In addition, no differences in effect were found between different types of blood products transfused. Conclusion: Blood transfusion was not associated with survival or time to recurrence after curative resection of PHC in this series. The alleged association is presumably related to the circumstances necessitating blood transfusions. Received 20 August 2015; accepted 31 August 2015

Correspondence T.M. van Gulik, AMC, Meibergdreef 9, 1105 AZ Amsterdam, The Netherlands. E-mail: t.m.vangulik @amc.uva.nl

Introduction Perihilar cholangiocarcinoma (PHC) is the most common type of cholangiocarcinoma,1 and originates at or near the biliary confluence. Surgical resection is the only potentially curative treatment for PHC, yielding a median overall survival of 19 to 39 months.2,3 Complete excision of PHC requires extended liver resection, which may cause significant blood loss. Perioperative blood transfusions with packed red blood cells (pRBC) are used to compensate for critical blood loss, but have been associated with increased risk of tumor recurrence and decreased long-term outcome in several tumor types, including colorectal, prostate, lung, and head and neck cancer.4–11 The effect of blood

HPB 2016, 18, 262–270

transfusions on prognosis is attributed to a distinct pathology of immunosuppression, known as transfusion-related immune modulation (TRIM). Blood transfusion seems to provoke an immune deviation12 and changes in the anti-inflammatory/proinflammatory environment.13–15 These substantial alterations form a complex and dynamic interplay creating a pro-tumor environment, which has been suggested to facilitate growth of residual cancer cells at the resection margin, transformation of micro-metastases into clinical metastases, or both. Despite multiple studies showing an association between perioperative blood transfusion and prognosis, it is unclear whether this effect is caused by clinical circumstances requiring transfusions or is due to the blood transfusion itself.16,17

© 2015 Published by Elsevier Ltd on behalf of International Hepato-Pancreato-Biliary Association Inc.

HPB

263

Conflicting results have been reported after resection of cholangiocarcinoma.18–23 Some studies included both proximal and distal cholangiocarcinoma, but these should be regarded as distinct tumor entities with different prognosis, as the latter involves the pancreatic head.24,25 The aim of this study was to assess the effect of perioperative blood transfusions on overall survival (OS) and time to recurrence in patients who recovered after resection of PHC, thus excluding patients who died from post-operative complications. As a secondary analysis, we also assessed the individual effects of different transfusion products.5

Methods A retrospective database was used, identifying 167 consecutive patients who underwent a curative-intent resection of PHC at a single center (Academic Medical Center, Amsterdam, The Netherlands) between 1992 and 2013. All patients who died within 90 days after resection (n = 22; 13.2%) were excluded. These patients most likely died from perioperative complications, which is a potential confounder when assessing the effect of transfusion on long-term outcome.26 PHC was defined as a pathologically confirmed biliary malignancy originating at the biliary confluence, right or left hepatic duct, or common hepatic duct.24 Patient selection and perioperative management have been described previously.27 Briefly, patients underwent routine preoperative biliary drainage, and preoperative low-dose radiotherapy (3 × 3.5 Gy) to prevent seeding metastases. An extrahepatic bile duct resection without liver resection was performed in patients with Bismuth type I tumors. For Bismuth type 2, 3 and 4 tumors, resection encompassed excision of the liver hilum en bloc with (extended) hemihepatectomy, excision of the portal vein bifurcation when involved, and complete lymphadenectomy of the hepatoduodenal ligament. Caudate lobectomy was performed in most patients since the late 90s.

Data collection and definitions Perioperative transfusion was defined as administration of one or more blood products within seven days before or after surgery. Blood transfusions were further classified into the different blood products administered, consisting of pRBC, fresh-frozen plasma (FFP) or platelets. Overall survival was measured from the date of surgery to the date of death. Patients were censored when alive at January 1st, 2014. Since survival status was synchronized with the Dutch municipal register, no patients were lost to follow-up of overall survival. Time to recurrence was measured from the time of surgery to the time of the first recurrence on imaging. No adjuvant chemotherapy was administered after initial curative resection. Clinical follow-up was performed routinely every three months in the first year after surgery and every six months in the following five years. Laboratory tests and follow-up CT scans were performed in the first six months to detect early recurrence, and in later course when indicated. Patients who had no

HPB 2016, 18, 262–270

observed recurrence were censored at the time of the last followup (not necessarily with imaging) prior to January 2014. Major complications were graded according to the Clavien–Dindo classification28; severe morbidity included grade III and IV complications (grade V, post-operative death, was excluded from the analysis).

Statistical analysis Continuous data are expressed as mean (±standard deviation) or median (±interquartile range (IQR)) as applicable. For comparing continuous variables a t-test or Mann–Whitney U test was used; for comparing proportions, Fisher’s exact test or Chi square test was used. Firstly, characteristics between patients with and without a blood transfusion were compared, including patient demographics, comorbidities, preoperative status and treatment, and perioperative details. Survival was analyzed using a Kaplan–Meier survival plot, and compared with a log rank test. All models in Cox multivariable survival analysis were adjusted for age, sex and known prognostic factors including resection margin, lymph node stage, tumor differentiation,29 and major complications.30 The same prognostic factors were used in the analysis assessing the impact of different blood products of OS. Statistical analysis was performed using SPSS (version 20, SPSS Inc., Chicago, IL). Two sided P-values < 0.05 were considered significant.

Results Patient characteristics Of 145 patients included, 80 (55.2%) received a perioperative blood transfusion: 26 patients (17.9%) received one or two blood units, and 54 patients (37.2%) received more than two blood units (total range, 1–47). Evaluation of different time spans shows a decrease in administered blood transfusions; 1992–1999 20 (62.5%), 2000–2006 25 (67.6%), 2007–2014 35 (46.1%). Table 1 details baseline and intra-operative characteristics of patients with and without perioperative transfusion. Patients receiving blood products were younger, more often jaundiced at presentation, and more often suffered from preoperative cholangitis. Also, the disease was more extensive in patients who had received perioperative blood transfusion, as evidenced by a higher Bismuth classification and Blumgart stage on preoperative imaging, and use of more extended resections and portal vein reconstruction. Patients with a perioperative blood transfusion more often had a major post-operative complication. Overall survival The median OS in the study cohort was 47 months (95% confidence interval [CI] 39–55). The median follow-up among survivors was 47 months (range 4–244). At last follow-up, 83 patients (57.2%) had died. Median OS of patients receiving a blood transfusion was 41 months compared to 49 months in

© 2015 Published by Elsevier Ltd on behalf of International Hepato-Pancreato-Biliary Association Inc.

HPB

264

Table 1 Baseline and intra-operative characteristics of patients with

and

without

perioperative

transfusion

for

perihilar

chol-

Table 1 (continued )

Variable

angiocarcinoma. Variables are expressed as numbers (count) and percentages, unless stated otherwise Variable

Age (years), mean (sd)

Transfusion PTotal No valuea (N [ 145) transfusion N [ 80 N [ 65 63 ± 11.1

66 ± 10.7

59 ± 10.7

Sex

0.108

Female

56

21 (37.5)

35 (62.5)

Male

89

44 (49.4)

45 (50.6)

ASA classification 126

54 (42.9)

72 (57.1)

3

16

8 (50.0)

8 (50.0)

Jaundice at presentation 30

18 (60.0)

12 (40.0)

Yes

112

46 (41.1)

66 (58.9)

Preoperative biliary drainage

0.074 19

11 (57.9)

5

1 (20.0)

4 (80.0)

Endoscopic

89

44 (49.4)

45 (50.6)

Both

32

9 (28.1)

23 (71.9)

Preoperative cholangitis

0.012

No

105

55 (52.4)

50 (47.6)

Yes

27

7 (25.9)

20 (74.1)

14 ± 8

11 ± 10

19 ± 28

0.488

8.0 ± 1.0

7.8 ± 1.1

0.197

Preoperative total bilirubin (mmol/L), median (IQR)

7.9 ± 1.1 Preoperative hemoglobin, mean (sd) Blumgart classification

0.002

1

70

40 (57.1)

30 (42.9)

2

29

7 (24.1)

22 (75.9)

3

22

6 (27.3)

16 (72.7)

Bismuth classification

<0.001

Left or right duct only

16

13 (81.2)

3 (18.8)

1/2

41

29 (70.7)

12 (29.3)

3A/B

69

17 (24.6)

52 (75.4)

4

19

6 (31.6)

13 (68.4)

No major hepatectomy

43

29 (67.4)

14 (32.6)

(Ext.) left hepatectomy

54

23 (42.6)

31 (57.4)

Type of resection

0.001

(continued on next column)

HPB 2016, 18, 262–270

35 (72.9)

Caudate lobe resection

0.129

No

65

33 (50.8)

32 (49.2)

Yes

80

32 (40.0)

48 (60.0) 0.023

No

118

58 (49.2)

Yes

27

7 (25.9)

60 (50.8) 20 (74.1) 31.8 ± 15.1

<0.001 0.002

Major post-operative complications (Clavien grade 3 or 4) No

94

51 (54.3)

43 (45.7)

Yes

50

14 (28.0)

36 (72.0)

N, number of patients.

8 (42.1)

Percutaneous

13 (27.1)

Blood loss (100 ml), 22.8 ± 15.4 12.5 ± 6.9 mean (sd) 0.050

No

48

PV reconstruction

0.389

1–2

No

0.001

(Ext.) right hepatectomy

Transfusion PTotal No valuea (N [ 145) transfusion N [ 80 N [ 65

patients without blood transfusions, which was not significant in univariable analysis (P = 0.46; Fig. 1a). To further explore a potential dose-dependent association, patients were stratified for the number of blood products transfused (none, 1 or 2, or 3). A trend towards worse OS was observed, but the effect remained non-significant (P = 0.23; Fig. 1b). Patients with three or more transfusions had a median OS of 39 months compared to a median OS of 41 months in patients who received one or two blood products. Multivariable analysis was performed to assess the association between one or more units transfused versus no units transfused. Risk-adjustment revealed significant associations between OS and known prognostic factors, including lymph node involvement, R1 resection, and tumor differentiation, but no association with blood transfusion (Table 2). Time to recurrence Recurrence status was unknown in 12 patients who were lost to follow-up of recurrences (8.3%). Among the other 133 patients, tumor recurrence was detected in 72 patients (54.1%) during follow up. Initial recurrences were classified according to the 7th edition of the AJCC staging system24 as local recurrence in 20 patients (15.0%), distant recurrence in 39 patients (29.3%), or both in 13 patients (9.8%). The median time to recurrence in all patients was 39 months; the median time to recurrence was 39 months among patients receiving a blood transfusion, and 44 months among patients without blood transfusions (P = 0.91; Fig. 2a). Similar to OS, stratification for the number of blood products (none, 1–2, or 3) revealed a trend in time to recurrence, but the effect was non-significant (P = 0.25; Fig. 2b). Interestingly, patients with one or two units transfused had a

© 2015 Published by Elsevier Ltd on behalf of International Hepato-Pancreato-Biliary Association Inc.

HPB

265

Figure 1 Kaplan-Meier curves for overall survival

Table 2 Prognostic factors for overall survival

Variable

Est. median survival (months)

Cox regression Univariable HR (95% CI)

P-value

Multivariable HR (95% CI)

P-value

Blood transfusion

No

49

Reference

0.461

Reference

0.993

Yes

41

1.18 (0.76–1.83)

Age

Per year

NA

1.00 (0.98–1.02)

0.651

0.99 (0.96–1.01)

0.150

Sex

Female

47

Reference

0.474

Reference

0.854

Male

47

1.18 (0.75–1.84)

Type of resection

No major hepatectomy

41

Reference

0.516

Reference

Major hepatectomy

48

0.86 (0.55–1.35)

No

46

Reference

Yes

NAa

0.51 (0.25–1.02)

R0

49

Reference

R1

38

1.46 (0.93–2.28)

Lymph node metastases

N0

49

Reference

N1

27

2.27 (1.38–3.73)

Tumor differentiation

Well

68

Reference

Moderate/poor

38

2.21 (1.22–4.02)

No

49

Reference

Yes

29

1.36 (0.86–2.15)

Portal vein reconstruction

Margin

Major complications

1.00 (0.59–1.68)

1.05 (0.63–1.74) 0.890

1.04 (0.61–1.76) 0.057

Reference

0.028

0.43 (0.20–0.91) 0.099

Reference

0.015

1.93 (1.14–3.28) 0.001

Reference

0.009

Reference

0.001

2.49 (1.17–4.28) 0.014

2.15 (1.17–3.96) 0.185

Reference

0.218

1.37 (0.83–2.28)

HR, hazard ratio; CI, confidence interval; NA, not available. The median survival in patients with PV reconstruction is not available, this is because more than half of the patients is alive. The median has not been reached.

a

non-significant longer time to recurrence compared to patients with no transfusions (median 61 versus 44 months, respectively). Time to recurrence was shortest in patients with three or more units transfused (median 33 months). Finally, the association between one or more transfused units (versus no units transfused) was assessed in multivariable analysis. Again, riskadjustment revealed significant associations of time to recurrence with known prognostic factors, including lymph node

HPB 2016, 18, 262–270

involvement, tumor differentiation and major complications but no association with one or more units transfused. (Table 3) Red blood cells, platelets concentrates and freshfrozen plasma Seventy-seven patients received pRBC (53.1%), 47 patients received FFP (32.4%) and 9 patients received platelet concentrates in (6.2%). Univariable analysis showed that FFP had a

© 2015 Published by Elsevier Ltd on behalf of International Hepato-Pancreato-Biliary Association Inc.

HPB

266

Figure 2 Kaplan-Meier curves for time to recurrence

Table 3 Prognostic factors for time to recurrence

Variable

Blood transfusion

No

Est. median time to recurrence (months)

Cox regression

44

Reference

Univariable HR (95% CI)

P-value 0.906

Multivariable HR (95% CI) Reference

P-value 0.989

Yes

39

0.97 (0.59–1.59)

Age

Per year

NA

0.99 (0.97–1.02)

0.528

1.00 (0.97–1.02)

0.636

Sex

Female

61

Reference

0.313

Reference

0.710

Male

39

1.30 (0.78–2.16)

No major hepatectomy

36

Reference

Major hepatectomy

44

0.75 (0.46–1.24)

No

40

Type of resection

Portal vein reconstruction

Margin

Lymph node metastasis

Tumor differentiation

Major complications

Reference a

Yes

NA

R0

44

Reference

R1

33

1.11 (0.65–1.90)

N0

68

Reference

N1

16

3.63 (2.13–6.20)

Well

1.00 (0.57–1.78)

1.12 (0.62–2.02) 0.263

Reference 33

2.49 (1.22–5.06)

No

46

Reference

Yes

27

1.51 (0.90–2.51)

0.656

0.88 (0.49–1.57) 0.610

Reference

0.703

Reference

0.61 (0.30–1.24)

Moderate/poor

Reference

0.354

0.69 (0.31–1.52) 0.392

1.31 (0.71–2.41) <0.001

Reference

<0.001

3.85 (2.13–6.96) 0.012

Reference

0.011

2.56 (1.24–5.31) 0.118

Reference

0.071

1.67 (0.96–2.91)

HR, hazard ratio; CI, confidence interval; NA, not available. The median survival in patients with PV reconstruction is not available, this is because more than half of the patients is alive. The median has not been reached.

a

non-significant adverse effect on OS (HR 1.36, 95% CI 0.87–2.12). After risk-adjustment in multivariable analysis, none of the three had a significant effect on OS. (Table 4)

Discussion We set out to assess the effect of perioperative blood transfusions on long-term outcomes after resection of perihilar

HPB 2016, 18, 262–270

cholangiocarcinoma in a single center, specialized in biliary tract surgery. After appropriate risk-adjustment, no effect of blood transfusion was found on tumor recurrence or overall survival. Blood transfusions have been associated with worse oncological outcomes in several types of cancer,9,16,17,31,32 including multiple studies in cholangiocarcinoma.21,22,33 These studies attributed this association to transfusion-related immune modulation (TRIM), which is characterized by an increase in T helper

© 2015 Published by Elsevier Ltd on behalf of International Hepato-Pancreato-Biliary Association Inc.

HPB

267

Table 4 Blood product-specific effect on overall survival

No. (%)

Univariable HR (95% CI)

P-value

Multivariable HR (95% CI)

P-value

Packed red blood cells

77 (53.1%)

1.18 (0.76–1.82)

0.470

0.88 (0.54–1.44)

0.601

Fresh-frozen plasma

47 (32.4%)

1.36 (0.87–2.12)

0.172

1.05 (0.66–1.68)

0.874

Platelet concentrates

9 (6.2%)

0.64 (0.23–1.75)

0.383

0.69 (0.25–1.95)

0.694

HR, hazard ratio; CI, confidence interval.

Table 5 Overview of recent studies assessing the effect blood transfusion on overall survival after resection of cholangiocarcinoma

Author, reference, publication year

N

Type of cholangiocarcinoma

Univariable analysis

Multivariable analysis

HR or RR (95% CI)

P-value

HR or RR (95% CI)

P-value

83

Perihilar

NA

0.01

2.00 (1.09–3.69)

0.03

574

Perihilar

NA

0.002

1.35 (1.06–1.72)

0.02

Any extrahepatic (hilar or distal)

2.05 (1.19–3.51)

0.01

1.14 (0.52–2.48)

0.75

Studies with significant association Young et al. 201121 Nagino et al. 2013

22

Studies with non-significant association Muller et al. 201441

128

Chauhan et al. 201118

51

Hilar

NA

0.37

–

Furusuwa et al. 201419

144

Hilar

NA

0.02

1.49 (0.90–2.47)

0.12

Intrahepatic

NA

0.05

1.98 (1.05–3.72)

0.44

Perihilar

1.18 (0.76–1.83)

0.46

1.00 (0.59–1.68)

0.99

Li et al. 201423 Present study

58a 145

N, number of patients in the study; HR, hazard ratio; RR, relative risk; NA, not available. Perioperative blood transfusion cut off was 600 mL.

a

type 2 (Th2) and regulatory T-cells. Th2 cytokines promote matrix metalloproteinase expression, invasiveness and metastasis of tumor cells.13 This in turn leads to down-regulation of the secretion of T helper type 1 cytokines, which normally suppress tumor growth and initiate changes to the inflammatory environment.12 Recent experimental evidence supports these findings.34–37 Taking this into account, in combination with other side effects such as transfusion-related acute lung injury (TRALI), current guidelines recommend a cautious policy to use perioperative blood transfusions only when it is highly required.38 The alleged association between blood transfusion and overall survival were not corroborated in more recent studies. Several studies found no effect of blood transfusion on overall survival,39,40 including several studies in cholangiocarcinoma (Table 5). Amongst others,18,19,23,41 Muller et al. found no effect of blood transfusions in a propensity-matched analysis of patients undergoing resection of distal or proximal extrahepatic cholangiocarcinoma.41 The present study adds weight to that observation by analyzing a large homogenous cohort restricted to patients undergoing resection of perihilar cholangiocarcinoma, which is associated with a higher risk of blood loss because of the extended liver resections required to achieve complete removal of the tumor. Moreover, we excluded all patients who died within 90 days after surgery, thereby eliminating a potential confounding bias of perioperative complications and

HPB 2016, 18, 262–270

truly focusing the analysis on long-term outcome. We found no association with time to tumor recurrence and no association with overall survival, neither in unadjusted nor in risk-adjusted analysis. In the abovementioned studies, the presumed effect of blood transfusion on long-term outcomes disappeared when the analysis was adjusted for the circumstances that require the blood transfusion. Additionally, one study showed no effect of blood transfusion in a large cohort of veterans undergoing surgery when the analysis was adjusted for perioperative complications.26 Other studies have suggested that anemia-induced hypoxia is the actual contributor to tumor recurrence instead of the blood transfusion.42–44 This theory was supported by the long-term results of a clinical trial comparing allogeneic versus autologous blood transfusion in colorectal cancer surgery.45 Counter-intuitively, this study showed worse overall and disease free survival after autologous transfusion, which was attributed to the induction of iatrogenic anemia before surgery. Furthermore, recent evidence demonstrates that a higher hemoglobin level mediates the response to radiation through delivery of oxygen to the tumor.46 These studies fit into a generally changing attitude towards perioperative blood transfusion: the assumption that perioperative blood transfusion is safe in terms of long-term outcome, and can be helpful when it is clinically required. A meta-analysis could provide a higher level of evidence to shed more light on this discussion. In that perspective, publication of

© 2015 Published by Elsevier Ltd on behalf of International Hepato-Pancreato-Biliary Association Inc.

HPB

268

negative studies is equally important as publishing positive studies, in order to prevent publication bias. In our study, a relatively large amount of patients (55%) received blood transfusion, reflecting a liberal transfusion policy during the study period. Some studies reported less frequent use of blood transfusions during resection of PHC,41,47 whereas other studies reported transfusion rates similar to the present study.18,19,39 The applied transfusion policy is apparently justified since no association of blood transfusion with survival has been found. Up to now, little is reported about the association between survival and fresh-frozen plasma or platelet concentrates,5,48 as literature mainly focuses on the effect of packed red blood cells.49,50 An interesting finding of our study is that the distinction between blood products has no influence on the outcome. Comparable results were reported for example by Tomimaru et al., who found that FFP transfusion did not affect cancer prognosis following hepatic resection for HCC.51 McGrath et al. concluded that platelet transfusions in cardiac surgery have no effect on morbidity after cardiac surgery.52 Several studies have shown a dose-dependent relationship, with three or more units of transfused blood almost doubling the risk observed with one or two units.9,17,53,54 Although, the principal analysis found no effect of blood transfusions on longterm outcomes, we cannot exclude a potential effect of massive blood transfusion, which could still cause immune modulation and alterations to the inflammatory response. Moreover, blood transfusion remains a costly therapy and unnecessary blood transfusions should be avoided. Transfusions can still cause alloimmunization, transmission of viral diseases, graft-versushost disease, and an increased post-operative infection rate.4,6,7,55 It is important to make an individualized decision on the use of blood transfusion in every patient undergoing cancer surgery, in conjunction with other precautionary measures.5 This retrospective study has several limitations. The cohort stretched over a long time period (1992–2013), which may have introduced heterogeneity related to changes in perioperative care. On the other hand, this approach enabled us to analyze the effect of blood transfusion in the largest cohort of resected PHC to date. Nevertheless, it remains a relatively small sample size with which a type II error cannot be ruled out. Furthermore, no information was available on the storage time of pRBC’s, which may also affect outcome. Limiting the study to perihilar cholangiocarcinoma improved homogeneity, as distal and intrahepatic cholangiocarcinoma are regarded as different tumor entities with their own specific treatment and prognosis. Although, we performed risk-adjustment for a variety of observed confounding factors, potential bias due to unknown or unobserved confounders cannot be excluded. In conclusion, perioperative blood transfusion was not associated with overall survival or time to recurrence in this large single-center cohort of patients who recovered from resection of perihilar cholangiocarcinoma. Therefore, the use of blood

HPB 2016, 18, 262–270

transfusion during or after surgery for perihilar cholangiocarcinoma seems safe in terms of long-term outcome. The alleged association of perioperative blood transfusions with worse outcomes after curative resection of PHC is presumably due to the circumstances necessitating blood transfusions instead of the blood transfusion per se. Funding sources None. Conflicts of interest None declared. References 1. Siegel R, Ma J, Zou Z, Jemal A. (2014) Cancer statistics, 2014. CA – Cancer J Clin 64:9–29. 2. Popescu I, Dumitrascu T. (2014) Curative-intent surgery for hilar cholangiocarcinoma: prognostic factors for clinical decision making. Langenbecks Arch Surg 399:693–705. 3. D’Angelica MI, Jarnagin WR, Blumgart LH. (2004) Resectable hilar cholangiocarcinoma: surgical treatment and long-term outcome. Surg Today 34:885–890. 4. Carson JL, Carless PA, Hebert PC. (2012) Transfusion thresholds and other strategies for guiding allogeneic red blood cell transfusion. Cochrane Database Syst Rev 4:CD002042. 5. Cata JP, Wang H, Gottumukkala V, Reuben J, Sessler DI. (2013) Inflammatory response, immunosuppression, and cancer recurrence after perioperative blood transfusions. Br J Anaesth 110:690–701. 6. Salpeter SR, Buckley JS, Chatterjee S. (2014) Impact of more restrictive blood transfusion strategies on clinical outcomes: a meta-analysis and systematic review. Am J Med 127:124–131. e3. 7. Theodoraki K, Markatou M, Rizos D, Fassoulaki A. (2014) The impact of two different transfusion strategies on patient immune response during major abdominal surgery: a preliminary report. J Immunol Res 2014: 945829. 8. Churchhouse AM, Mathews TJ, McBride OM, Dunning J. (2012) Does blood transfusion increase the chance of recurrence in patients undergoing surgery for lung cancer? Interact Cardiovasc Thorac Surg 14: 85–90. 9. Gunka I, Dostalik J, Martinek L, Gunkova P, Mazur M. (2013) Impact of blood transfusions on survival and recurrence in colorectal cancer surgery. Indian J Surg 75:94–101. 10. Miki C, Hiro J, Ojima E, Inoue Y, Mohri Y, Kusunoki M. (2006) Perioperative allogeneic blood transfusion, the related cytokine response and long-term survival after potentially curative resection of colorectal cancer. Clin Oncol 18:60–66. 11. Luan H, Ye F, Wu L, Zhou Y, Jiang J. (2014) Perioperative blood transfusion adversely affects prognosis after resection of lung cancer: a systematic review and a meta-analysis. BMC Surg 14:34. 12. Vamvakas EC. (2002) Possible mechanisms of allogeneic blood transfusion-associated postoperative infection. Transfus Med Rev 16:144–160. 13. Vamvakas EC, Blajchman MA. (2007) Transfusion-related immunomodulation (TRIM): an update. Blood Rev 21:327–348. 14. Lacy AM, Garcia-Valdecasas JC, Delgado S, Castells A, Taura P, Pique JM et al. (2002) Laparoscopy-assisted colectomy versus open colectomy for treatment of non-metastatic colon cancer: a randomised trial. Lancet 359:2224–2229.

© 2015 Published by Elsevier Ltd on behalf of International Hepato-Pancreato-Biliary Association Inc.

HPB

269

15. Bouvy ND, Marquet RL, Jeekel H, Bonjer HJ. (1996) Impact of gas(less)

31. Cata JP, Chukka V, Wang H, Feng L, Gottumukkala V, Martinez F et al.

laparoscopy and laparotomy on peritoneal tumor growth and abdominal

(2013) Perioperative blood transfusions and survival in patients with

wall metastases. Ann Surg 224:694–700. discussion-1.

non-small cell lung cancer: a retrospective study. BMC Anesthesiol 13:

16. Busch OR, Hop WC, Hoynck van Papendrecht MA, Marquet RL, Jeekel J. (1993) Blood transfusions and prognosis in colorectal cancer. N Engl J Med 328:1372–1376.

42. 32. Han DH, Choi GH, Park JY, Ahn SH, Kim KS, Choi JS et al. (2014) Lesson from 610 liver resections of hepatocellular carcinoma in a single

17. Amato A, Pescatori M. (2006) Perioperative blood transfusions for the recurrence of colorectal cancer. Cochrane Database Syst Rev, CD005033.

center over 10 years. World J Surg Oncol 12:192. 33. Kimura N, Toyoki Y, Ishido K, Kudo D, Yakoshi Y, Tsutsumi S et al. (2015 Jun) Perioperative blood transfusion as a Poor prognostic factor

18. Chauhan A, House MG, Pitt HA, Nakeeb A, Howard TJ, Zyromski NJ et al. (2011) Post-operative morbidity results in decreased long-term

after Aggressive surgical resection for hilar cholangiocarcinoma. J Gastrointest Surg – Off J Soc Surg Aliment Tract 19:1194–1195.

survival after resection for hilar cholangiocarcinoma. HPB 13:139–147.

34. Ghio M, Contini P, Negrini S, Mazzei C, Zocchi MR, Poggi A. (2011)

19. Furusawa N, Kobayashi A, Yokoyama T, Shimizu A, Motoyama H,

Down regulation of human natural killer cell-mediated cytolysis induced

Miyagawa S. (2014) Surgical treatment of 144 cases of hilar chol-

by blood transfusion: role of transforming growth factor-beta(1), soluble

angiocarcinoma without liver-related mortality. World J Surg 38:

Fas ligand, and soluble Class I human leukocyte antigen. Transfusion 51:1567–1573.

1164–1176. 20. de Boer MT, Molenaar IQ, Porte RJ. (2007) Impact of blood loss on outcome after liver resection. Dig Surg 24:259–264.

35. Baumgartner JM, Nydam TL, Clarke JH, Banerjee A, Silliman CC, McCarter MD. (2009) Red blood cell supernatant potentiates LPS-

21. Young AL, Igami T, Senda Y, Adair R, Farid S, Toogood GJ et al. (2011)

induced proinflammatory cytokine response from peripheral blood

Evolution of the surgical management of perihilar cholangiocarcinoma in

mononuclear cells. J Interferon Cytokine Res – Off J Int Soc Interferon

a Western centre demonstrates improved survival with endoscopic

Cytokine Res 29:333–338.

biliary drainage and reduced use of blood transfusion. HPB 13:483–493.

36. Upile T, Jerjes W, Mahil J, Upile N, Sudhoff H, Wright A et al. (2011) An

22. Nagino M, Ebata T, Yokoyama Y, Igami T, Sugawara G, Takahashi Y

explanation for the worsened prognosis in some cancer patients of

et al. (2013) Evolution of surgical treatment for perihilar chol-

perioperative transfusion: the time-dependent release of biologically

angiocarcinoma: a single-center 34-year review of 574 consecutive

active growth factors from stored blood products. Eur Arch Oto-RhinoLaryngol – Off J Eur Fed Oto-Rhino-Laryngol Soc 268:1789–1794.

resections. Ann Surg 258:129–140. 23. Li H, Wu JS, Wang XT, Lv P, Gong LS, Liu G et al. (2014) Factors

37. Long K, Meier C, Ward M, Williams D, Woodward J, Bernard A. (2013)

predicting surgical resection in patients with intrahepatic chol-

Immunologic profiles of red blood cells using in vitro models of trans-

angiocarcinoma and cirrhosis. J Investig Surg – Off J Acad Surg Res 27:

fusion. J Surg Res 184:567–571. 38. Ramsey G, Wagar EA, Grimm EE, Friedberg RC, Souers RJ,

219–225. 24. Edge SB, Compton CC. (2010) The American Joint Committee on

Lehman CM. (2015) Red blood cell transfusion practices: a college of

cancer: the 7th edition of the AJCC cancer staging manual and the

American pathologists q-probes study of compliance with audit criteria in 128 hospitals. Arch Pathol Lab Med 139:351–355.

future of TNM. Ann Surg Oncol 17:1471–1474. 25. van der Gaag NA, Kloek JJ, de Bakker JK, Musters B, Geskus RB,

39. Warschkow R, Guller U, Koberle D, Muller SA, Steffen T, Thurnheer M

Busch OR et al. (2012) Survival analysis and prognostic nomogram for

et al. (2014) Perioperative blood transfusions do not impact overall and

patients undergoing resection of extrahepatic cholangiocarcinoma. Ann

disease-free survival after curative rectal cancer resection: a propensity

Oncol – Off J Eur Soc Med Oncol/ESMO 23:2642–2649.

score analysis. Ann Surg 259:131–138.

26. Lee J, Radulescue V, Porhomayon J, Pourafkari L, Arora P,

40. Pang TC, Spiro C, Ramacciotti T, Choi J, Drummond M, Sweeney E

Dosluoglu HH et al. (2015 Jan) The role of perioperative transfusion on

et al. (2015 Feb) Complications following liver resection for colorectal

long-term survival of veterans undergoing surgery. Ann Surg 261: 104–110.

metastases do not impact on longterm outcome. HPB 17:185–193. 41. Muller SA, Mehrabi A, Rahbari NN, Warschkow R, Elbers H, Leowardi C

27. van Gulik TM, Kloek JJ, Ruys AT, Busch OR, van Tienhoven GJ,

et al. (2014) Allogeneic blood transfusion does not affect outcome after

Lameris JS et al. (2011) Multidisciplinary management of hilar chol-

curative resection for advanced cholangiocarcinoma. Ann Surg Oncol

angiocarcinoma (Klatskin tumor): extended resection is associated with improved survival. Eur J Surg Oncol – J Eur Soc Surg Oncol Br Assoc

21:155–164. 42. Vaupel P, Mayer A. (2005) Hypoxia and anemia: effects on tumor biology and treatment resistance. Transfus Clin Biol – J de la Soc

Surg Oncol 37:65–71. 28. Dindo D, Demartines N, Clavien PA. (2004) Classification of surgical complications: a new proposal with evaluation in a cohort of 6336 patients and results of a survey. Ann Surg 240:205–213. 29. Groot Koerkamp B, Wiggers JK, Gonen M, Doussot A, Allen PJ, Besselink MG et al. (2015 Sep) Survival after resection of perihilar

Francaise de Transfus Sang 12:5–10. 43. Leo C, Giaccia AJ, Denko NC. (2004) The hypoxic tumor microenvironment and gene expression. Semin Radiat Oncol 14:207–214. 44. Bindra RS, Crosby ME, Glazer PM. (2007) Regulation of DNA repair in hypoxic cancer cells. Cancer Metastasis Rev 26:249–260.

cholangiocarcinoma-development and external validation of a prog-

45. Harlaar JJ, Gosselink MP, Hop WC, Lange JF, Busch OR, Jeekel H. (2012)

nostic nomogram. Ann Oncol – Off J Eur Soc Med Oncol/ESMO 26:

Blood transfusions and prognosis in colorectal cancer: long-term results of a randomized controlled trial. Ann Surg 256:681–686. discussion 6–7.

1930–1935. 30. Pucher PH, Aggarwal R, Qurashi M, Darzi A. (2014) Meta-analysis of the

46. Katahira-Suzuki R, Hata M, Tateishi U, Taguchi T, Takano S, Omura-

effect of postoperative in-hospital morbidity on long-term patient sur-

Minamisawa M et al. (2015) Definitive chemo-radiotherapy for squa-

vival. Br J Surg 101:1499–1508.

mous cell carcinoma of the pharynx: impact of baseline low hemoglobin

HPB 2016, 18, 262–270

© 2015 Published by Elsevier Ltd on behalf of International Hepato-Pancreato-Biliary Association Inc.

HPB

270

level (<12 g/dL) and post-radiation therapy F-18 FDG-PET/CT. Ann Nucl

following hepatic resection for hepatocellular carcinoma. World J

Med 29:37–45.

Gastroenterol – WJG 16:5603–5610.

47. Liu CL, Fan ST, Lo CM, Tso WK, Lam CM, Wong J. (2006) Improved

52. McGrath T, Koch CG, Xu M, Li L, Mihaljevic T, Figueroa P et al. (2008)

operative and survival outcomes of surgical treatment for hilar chol-

Platelet transfusion in cardiac surgery does not confer increased risk for

angiocarcinoma. Br J Surg 93:1488–1494.

adverse morbid outcomes. Ann Thorac Surg 86:543–553.

48. Spiess BD, Royston D, Levy JH, Fitch J, Dietrich W, Body S et al. (2004)

53. Yao HS, Wang Q, Wang WJ, Hu ZQ. (2008) Intraoperative allogeneic red

Platelet transfusions during coronary artery bypass graft surgery are

blood cell transfusion in ampullary cancer outcome after curative

associated with serious adverse outcomes. Transfusion 44:1143–1148.

pancreatoduodenectomy: a clinical study and meta-analysis. World J

49. Cohen B, Matot I. (2013) Aged erythrocytes: a fine wine or sour grapes? Br J Anaesth 111(Suppl. 1):i62–70. 50. Bernard

AC,

Davenport

DL,

Surg 32:2038–2046. 54. Philips P, Farmer RW, Scoggins CR, McMasters KM, Martin, RC, 2nd.

TB,

(2013) Caudate lobe resections: a single-center experience and

Zwischenberger JB. (2009) Intraoperative transfusion of 1 U to 2 U

Chang

PK,

Vaughan

evaluation of factors predictive of outcomes. World J Surg Oncol 11:

packed red blood cells is associated with increased 30-day mortality, surgical-site infection, pneumonia, and sepsis in general surgery patients. J Am Coll Surg 208, 931–7, 7 e1–2; discussion 8–9. 51. Tomimaru Y, Wada H, Marubashi S, Kobayashi S, Eguchi H, Takeda Y

220. 55. Rohde JM, Dimcheff DE, Blumberg N, Saint S, Langa KM, Kuhn L et al. (2014) Health care-associated infection after red blood cell transfusion: a systematic review and meta-analysis. JAMA 311:1317–1326.

et al. (2010) Fresh frozen plasma transfusion does not affect outcomes

HPB 2016, 18, 262–270

© 2015 Published by Elsevier Ltd on behalf of International Hepato-Pancreato-Biliary Association Inc.